International Journal of Obesity (****) **, **** *556
& 2006 Nature Publishing Group All rights reserved 0307-0565/06 $30.00
www.nature.com/ijo
ORIGINAL ARTICLE
Multinutrient supplement containing ephedra and
caffeine causes weight loss and improves metabolic
risk factors in obese women: a randomized controlled
trial
RM Hackman1, PJ Havel1, HJ Schwartz2, JC Rutledge2, MR Watnik3, EM Noceti1, SJ Stohs4,
JS Stern1,2 and CL Keen1,2
1
Department of Nutrition, University of California, Davis, CA, USA; 2Department of Internal Medicine, University of
California, Davis, CA, USA; 3Department of Statistics, California State University, East Bay, CA, USA and 4Department of
Pharmacy Science, Creighton University, Omaha, USA
Objective: To determine the safety and efficacy of a dietary supplement with a low dose of ephedra and caffeine in overweight/
obese premenopausal female subjects.
Design: A 9-month, double-blind, randomized control study compared the efficacy and safety of a dietary supplement with
ephedra and caffeine to a control supplement.
Subjects: Sixty-one healthy, premenopausal women with body mass index (BMI) from 27 to 39 kg/m2 were randomly assigned
and received a dietary supplement (40 mg/day ephedra alkaloids, 100 mg/day caffeine, high potency mixture of vitamins,
minerals, omega-3 fatty acids) or a control supplement for 9 months.
Measurements: Efficacy: changes in body weight, body composition, lipids, insulin, leptin, adiponectin, ghrelin, and self-reports
of physical activity, diet and quality of life indices. Safety: blood pressure, heart rate, electrocardiograms, urinalysis, blood
histology, serum chemistry measures and self-reported symptoms.
Results: Forty-one women completed the study. The treatment group lost significantly more body weight ( 7.18 kg) and body
fat ( 5.33 kg) than the control group ( 2.25 and 0.99 kg, respectively), and showed significant declines in heart rate, serum
cholesterol, triglycerides, cholesterol to high-density lipoprotein ratio, glucose, fasting insulin, and leptin. Blood pressure,
electrocardiograms, other clinical chemistry measures, blood histology, urinalysis, and self-reported physical activity were similar
in the groups. Minor symptoms included dry mouth, insomnia, nervousness and palpitations. The treatment group reported
more energy and decreased appetite compared to controls and scored higher on a quality of life domain assessing vitality.
Conclusion: A dietary supplement containing a low potency ephedra/caffeine mixture appeared safe and effective in causing
loss of weight and body fat, and improving several metabolic parameters, including insulin sensitivity and lipid profiles when
tested under physician supervision. Such supplements could be a useful tool to assist with weight loss.
International Journal of Obesity (2006) 30, 1545 1556. doi:10.1038/sj.ijo.0803283; published online 21 March 2006
Keywords: (MeSH) ephedra; weight loss; dietary supplement; lipids; insulin; leptin
contributes to at least 300 000 deaths per year,2 and
Introduction
conservatively costs an estimated $75 billion annually in
health care and related costs in the US.3
An estimated 64% of American adults are overweight or
obese (body mass index (BMI) in kg/m2X25.0).1 Obesity is Physicians tend to underreport obesity, and less than half
of obese persons are advised to lose weight or offered a
associated with a number of chronic health problems,
supervised diet or exercise program by their doctor.4,5 These
trends may encourage people seeking weight loss to respond
Correspondence: Dr RM Hackman, Department of Nutrition, University of to heavily marketed products such as nutrition supplements,
California, One Shields Avenue, Davis, CA 95616-8669, USA.
many of which have limited evidence of safety or efficacy.
E-mail: abo92q@r.postjobfree.com
Studies in the early 1990s found that a combination of
Received 24 June 2005; revised 6 January 2006; accepted 19 January 2006;
ephedrine and caffeine significantly reduced body weight
published online 21 March 2006
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1546
and fat in humans over a 6-month period.6 Subsequent Table 1 Baseline characteristics of the subjectsa
2- and 6-month studies using combinations of ephedra (Ma
Control (N 32) Treatment (N 29)
Characteristic
Huang) and caffeine reported similar weight loss findings.7,8
The US Food and Drug Administration (FDA) banned the sale Race or ethnic group
African American 1 2
of dietary supplements containing ephedrine alkaloids in
Filipino American 2 0
2004, citing concerns such as consumption of high, Latino American 2 6
unregulated dosages, and use by at-risk populations with Mexican American 0 1
comorbid conditions.9 However, in 2005, a US District Court Asian 1 1
Polynesian 1 0
found insufficient evidence of adverse events related to the
Caucasian 24 19
sale of a dietary supplement containing 10 mg/day ephedra Did not reply 1 0
alkaloids, and ordered the FDA to rewrite its ephedra rule
35.570.9 38.471.1
with consideration for dose response relationships (Nutra- Age (year)
31.670.5 32.470.6
Body mass indexb
ceutical Corp. and Solaray, Inc. vs Lester Crawford, Acting
84.371.7 88.672.4c
Weight (kg)
Commissioner, US FDA et al.). 10571.4 10771.7
Systolic blood pressure (mm Hg)
Intake of vitamins, minerals, and omega-3 fatty acids may 6871.2 7071.3
Diastolic blood pressure (mm Hg)
69.571.5 67.171.4
be compromised in overweight and obese individuals due to Heart rate
improper dietary choices, and a multinutrient supplement
Cholesterol (mg/dl)
may be valuable in providing basic micronutrient support. 197.476.4 182.475.8
Total
We tested the hypothesis that a high-potency multivitamin 124.975.7 113.775.1
Low-density lipoprotein
48.571.9 47.272.1
and mineral formula, plus additional omega-3 fatty acids High-density lipoprotein
105.8711.7 107.8711.0
Triglycerides (mg/dl)
and botanical extracts including a low level of ephedra
alkaloids and caffeine would have a greater effect in Plus-minus values are means7s.e.m. bBody mass index is the weight in
a
kilograms divided by the square of the height in meters. cA significant
reducing body weight and body fat than a control formula
difference (Po0.05) in body weight existed between the two groups after
containing a lower potency multivitamin and mineral
randomization at baseline.
formula devoid of omega-3 fatty acids, botanical extracts,
ephedra alkaloids, and caffeine over a 9-month period. We
further evaluated the effects of these two types of supple-
Study design
ments on cardiovascular and metabolic indices, serum
chemistry, self-reported symptoms, and behavioral and Randomization of an equal proportion of subjects to one of
psychosocial measures. two groups was conducted with a random number generator
in multiples of 20.10 Subjects and research staff were blinded
to group assignment throughout the intervention. A sealed
Methods copy of the code was available to the study physicians for
emergency purposes. The statistician was the only person
Subjects with access to the code until completion of the study and
Sixty-one healthy pre-menopausal women, aged 25 47 years data verification.
old, with BMI from 27 to 39 kg/m2 were enrolled in the study The control group received a supplement containing
(Table 1). During an enrollment period from February to vitamins and minerals at 100% Daily Value, a small amount
August 2002, 216 prospective volunteers responded to of lutein, a corn oil capsule, and a cellulose tablet (Table 2).
recruitment. Ninety-four passed a telephone-based health The treatment group received a high-potency multivitamin
screening and were further evaluated by a health interview, and mineral formula, an omega-3 fatty acid capsule and a
physical exam, and clinical chemistry. Exclusion criteria botanical supplement containing Garcinia cambogia extract,
included resting systolic blood pressure greater than green tea extract, ephedra (Ephedra sinica) extract, guarana
140 mm Hg or diastolic blood pressure greater than (Paullinia cupana) extract (a source of caffeine), and other
90 mm Hg, history of, or existence of, any medical condition, botanical extracts (Table 2). The treatment group received a
total of 40 mg of ephedra alkaloids per day, provided as
use of prescription medications (except birth control pills),
use of antihistamines or other medications used for mild 20 mg prior to breakfast and 20 mg prior to lunch, and
asthma, coughs, colds, and allergies for 30 days prior to 100 mg of caffeine per day, provided as 50 mg prior to
randomization and for the duration of the study, caffeine breakfast and 50 mg prior to lunch. Independent analysis
intake greater than 150 mg/day, smoking presently or at least conducted at the Department of Pharmacy Science, Creight-
6 months prior to the study, involvement in any weight loss on University, confirmed the ephedrine and caffeine con-
program or diet presently or at least 6 months prior to the tent. Study supplements were provided in daily strip packs
study, and women who were pregnant, lactating or planning and were of similar size, shape and color. Subjects were
a pregnancy during the study period. All subjects provided instructed to take the botanical caplets or placebo 30 45 min
written informed consent, and the University of California, prior to breakfast and again prior to lunch, and to take the
multivitamin-mineral caplets and omega-3 fatty acid or corn
Davis Institutional Review Board approved the study protocol.
International Journal of Obesity
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1547
oil capsule with either lunch or dinner. Compliance was
Table 2 Composition of supplements for control and treatment groups
assessed at each clinic visit and by scheduled telephone
Ingredient U/day DRI Amount
interviews weekly, then monthly.
Women were instructed to refrain from eating or drinking
Control Treatment
beverages containing caffeine or alcohol for at least 12 h, and
a
Vitamin A (palmitate) mg 700-****-***
to refrain from heavy physical activity for 3 h, prior to the
Vitamin A (beta-carotene)a mg 700-***-****
clinic visits. Normal hydration was maintained during the
Vitamin C mg 75 60 1200
Vitamin Db mg 5 10 10 testing. Body weight was measured to the nearest 0.1 kg
Vitamin Ec mg 15 13.5 70.5
while wearing light clothing and without shoes using an
Vitamin K mg 90 25 0
electronic scale (Scale-Tronix 6002). Height was measured to
Thiamin mg 1.1 1.5 9
the nearest 0.5 cm without shoes using a wall-mounted
Riboflavin mg 1.1 1.7 10.2
Niacind mg 14 20 160 stadiometer (Ayrton S100). Body composition was assessed
Vitamin B6 mg 1.3 2 16
with a Xitron Hydra 4200 bioelectrical impedance spectro-
Folatee mg 400 400 800
meter (Xitron Technologies, San Diego, CA, USA) in a supine
Vitamin B12 mg 2.4 6 36
position 30 min after voiding.11 The manufacturer s software
Biotin mg 30 30 100
Pantothenic Acid mg 5 10 40 was used to calculate body composition. Body weight and
Calcium mg 100*-***-*** body composition were measured at baseline and at months
Iron mg 18 18 0
1.5, 3, 6, and 9 after beginning the supplement regimens.
Phosphorus mg 700 109 100
With the subject remaining supine after measurement of
Iodine mg 150 150 0
Magnesium mg 320 100 200 body composition, a 12-lead resting electrocardiograph
Zinc mg 8 15 15 (ECG) (Welch Allyn) was recorded and subsequently read
Selenium mg 55 20 80
by a board-certified cardiologist (JCR). Electrocardiograph s
Copper mg 900-****-****
were recorded at baseline, weeks 1, 2, and 4, and at months
Manganese mg 1.8 2 4
Chromium mg 25 120 200 1.5, 3, 6, and 9.
Molybdenum mg 45 75 50
Blood pressure was measured three times in a resting,
f
Chloride mg 72 0
seated position with an aneroid sphygmomanometer (Welch
f
Potassium mg 80 300
Allyn), according to standardized methods.12 A 1-min rest
f
Boron mg 150 300
f
Nickel mg 5 0 period was observed between each measurement. The last
f
Silicon mg 2 500
two readings were averaged for the final measurement.
f
Tin mg 10 0
Heart rate was measured in beats per minute after the second
f
Vanadium mg 10 50
blood pressure reading. Blood pressure and heart rates
f
Lutein mg 250 0
Eicosapentaenoic acid mg 0 0 180 were recorded at baseline, weeks 1, 2, and 4, and monthly
Docosahexaenoic acid mg 0 0 120 thereafter. Accepted values for normal blood pressure at the
Choline mg 0 0 60
time of the study (130/85 mm Hg) were used.13
Coenzyme Q-10 mg 0 0 150
Blood collected from the antecubital vein and urine
Garlic extract mg 0 0 50
L-Glutathione mg 0 0 5 samples were obtained at baseline and at months 1.5, 3, 6,
L-Methionine mg 0 0 100 and 9. Serum levels of sodium, potassium, bicarbonate,
Bioflavonoids mg 0 0 100
chloride, magnesium, carbon dioxide, urea nitrogen, creati-
L-Carnitine mg 0 0 25
nine, glucose, phosphorus, calcium, protein, albumin, alka-
f
Taurine mg 0 50
f
Beta-sitosterol mg 0 25 line phosphatase, aspartate transaminase (AST/SGOT), total
f
Gingko biloba extract mg 0 10
bilirubin, alanine transaminase (ALT/SGPT), creatinine
f
Silymarin extract mg 0 5
kinase, lactate dehydrogenase, ketone bodies, cholesterol,
f
Red wine polyphenols mg 0 5
high-density lipoprotein (HDL) cholesterol, and triglycerides
f
Garcinia cambogia extract mg 0 2000
f
Tulsi extract mg 0 50 were determined by chemistry analyzers (Beckman LXI
f
Gymnema sylvestre extract mg 0 10
and LX20 Pro). Low-density lipoprotein (LDL) cholesterol
f
Green tea extract mg 0 20
was calculated (TChol HDL TG/5 Calculated LDL). Insulin,
f
Oolong tea extract mg 0 200
thyroxine, and thyroid stimulating hormone were assessed
f
Eleutherococcus senticosis extract mg 0 50
f
550g
Guarana extract mg 0 by chemiluminescent immunoassay (Bayer Advia Centaur).
f
500h
Ephedra extract mg 0
The above tests, as well as blood histology, were performed
by the Department of Pathology at the University of
a
As retinol activity equivalents. bAs cholecalciferol. cAs alpha-tocopherol. dAs
niacin equivalents. eAs dietary folate equivalents. fNo DRI established. California Davis Medical Center. Urinalyses by dipstick and
g
Contains 100 mg caffeine. hContains 40 mg ephedrine group alkaloids. urine pregnancy tests were performed at our clinic.
Plasma glucose concentrations were measured with a
YSI Glucose Analyzer Model STAT 2300 (Yellow Springs
Instruments, Yellow Springs, OH, USA). Free fatty acids
were assayed with an enzymatic colorimetric assay (Wako
International Journal of Obesity
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1548
Chemicals USA, Richmond, VA, USA) adapted to a microtiter response terms (e.g. blood pressure above normal values).
plate reader. Intra- and interassay coefficients of variation in our Reported P-values were two-sided and a P-value of 0.05 or
laboratory are 4.7 and 7.5%, respectively. Insulin was assayed less was considered statistically significant. Analyses were
as described elsewhere,14 using human insulin standards performed with SAS software (version 9.1).23
(Linco Research, St Charles, MO, USA), (3-125I) insulin
(human recombinant) (Amersham Biosciences, Piscataway,
NJ, USA) and antiinsulin antisera (Radioassay Systems Results
Laboratories, Carson, CA, USA). Intra- and interassay varia-
tions are 5.5 and 8.9%, respectively. Leptin was measured Attrition
with a radioimmunoassay kit using an I-125-iodinated Of the 61 subjects enrolled, 41 (67%) completed the 9-
human leptin tracer and human leptin standards (Linco month study. Figure 1 depicts the randomization flow and
Research, St Charles, MO, USA). Intra- and interassay reasons for attrition. The study code never needed to be
variations are 5.8 and 5.7%, respectively. Adiponectin was broken for emergency purposes. No subjects were removed
determined with a radioimmunoassay kit using an I-125- from the study due to clinically abnormal changes in serum
iodinated murine adiponectin tracer, a multispecies adipo- chemistry, hematology, urinalysis, pregnancy, or ECG read-
nectin rabbit antiserum, and human adiponectin standards ings. No subjects were removed by the study physicians due
(Linco Research, St Charles, MO, USA). Intra- and interassay to severe or serious symptoms. Of the nine subjects in the
variations are 5.1 and 8.6%, respectively. Ghrelin was treatment group who withdrew, two started a medication
measured with a radioimmunoassay kit using rabbit anti- that was exclusionary for the study, two reported headaches,
serum specific for the synthetic peptide (Phoenix Pharma- one was lost to follow-up, one was noncompliant and one
ceuticals, Belmont, CA, USA). Intra- and interassay variations each reported insomnia, nervousness, and dizziness. Of the
are 5.9 and 13.3%, respectively. The homeostasis assessment nine subjects in the control group who withdrew, four were
model (HOMA-IR) was used to calculate an index of insulin lost to follow-up, three reported personal conflicts, one
resistance/sensitivity.15 refused to participate, and one started a medication that was
Self-reported symptoms were assessed in person at base- contraindicated for study participation.
line, weeks 1, 2, and 4 and monthly hereafter, as well as by
regularly scheduled telephone calls. Events were self-rated
for intensity (mild, moderate, severe, or serious), reviewed Body weight and composition
by the study physicians and classified according to the The treatment group had a slightly higher average body
FDA coding system and thesaurus for adverse events weight (89.0 kg) than the control group (84.0 kg) at baseline.
terminology.16 Body weights were rank transformed prior to statistical
Food intake was assessed by a standardized self-adminis- analysis, and all values from subjects completing 3 months
tered food frequency questionnaire (Block Food Frequency of the intervention were included in the model. Significant
Questionnaire17) at baseline and month 9. Physical activity differences between groups for weight loss occurred at
months 1.5, 3, 6, and 9 (Po0.0001 between groups at each
was assessed by an interview-administered standardized
instrument (Seven-Day Physical Activity Recall18) at baseline time; Figure 2), with the treatment group losing significantly
and at months 1.5, 3, 6, and 9. Psychosocial indices were more body weight than the control group (overall change
determined at baseline and months 1.5, 3, 6, and 9 using a between groups, P 0.0022). Changes in BMI were similar to
self-administered, standardized survey (Medical Outcomes those found for changes in body weight, with the treatment
Study 36-item Health Survey [SF-36 version 2]19) to assess group showing a significant reduction in BMI at months 1.5,
3, 6, and 9 compared to controls (Po0.0001 between groups
domains of general health status, limitations in physical
activities due to health problems, limitations in usual role at each time; P 0.0002 overall). The amount of body fat did
activities due to physical or emotional problems, bodily not differ between groups at baseline (P 0.71). Significant
pain, energy and fatigue levels, social functioning, general reductions in body fat were noted for the treatment group
mental health and general health perception. at months 1.5, 3, 6, and 9 compared to controls at each
time, and was significant overall (P 0.0017; Figure 2). No
differences in fat-free mass were found between the two
Statistical analysis groups at specific visits (P 0.60).
Outcome measures were analyzed using mixed model
analysis of variance. Least squares means comparisons were
Blood pressure, heart rate, and electrocardiograph
performed to determine significant differences between
groups overall and between the groups at specific time Mean systolic and diastolic blood pressure values remained
points. When model assumptions were not met, transforma- fairly constant throughout the intervention and were within
tions were used prior to analysis.20 Missing values were not the normal range at all measurement periods. At the time of
imputed; they were left as missing.21,22 Fisher s exact test was the study, normal blood pressure was defined as systolic
o140 mm Hg and diastolic o90 mm Hg. Using these criteria,
used to determine differences between groups for categorical
International Journal of Obesity
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1549
217 Assessed for Eligibility
123 Excluded
86 Not Meeting Inclusion Criteria
37 Refused to Participate
94 Medically Screened
33 Excluded
33 Not Meeting Inclusion Criteria
61 Randomized
1 Excluded
1 Refused to Participate
29 Assigned to 31 Assigned to
Treatment Group Control Group
3 Excluded
2 Lost to Follow-up
1 Non-compliance
1 Insomnia
1 Nervousness
26 at 6 weeks 29 at 6 weeks
4 Excluded 6 Excluded
1 Lost to Follow-up 2 Lost to Follow-up
1 Dizziness 3 Personal Conflict
1 Headaches 1 Medication
1 Medication
22 at 12 weeks 23 at 12 weeks
2 Excluded
1 Medication
1 Headaches
20 at 24 weeks 23 at 24 weeks
1 Excluded
1 Surgery
23 at 36 weeks
19 at 36 weeks
Figure 1 Randomization flow.
no differences in the number of high values for systolic or were noted for a small number of subjects in each group
diastolic blood pressure were found between groups. By the at various times, but in no case were the readings suffi-
completion of the study, new standards had been adopted to ciently alarming to warrant discontinuation of any subject.
define normal blood pressure as systolic o120 mm Hg and One subject in the treatment group was found to have a
diastolic o80 mm Hg. Using these newer criteria, no differ- heart murmur, confirmed by an echocardiogram, which was
ences in systolic blood pressure were observed between diagnosed by the study cardiologist during the blinded phase
groups except at month 3. No differences in diastolic values of the intervention. The murmur most likely was due to a
were found (Table 3). congenital anomaly and was not considered to be clinically
Heart rates were similar between groups at baseline and significant.
were log transformed for analysis. Following ANOVA, mean
values and 95% confidence intervals were calculated24 and
Serum lipids, chemistry, urinalysis, and histology
are shown in Table 3. Heart rates in the control group
were significantly higher at week 1, month 3, and month 9 Total serum cholesterol levels were significantly higher in
compared to the treatment group. Heart rates for the the control group at baseline than in the treatment group
(P 0.002). Both groups showed a similar decline in
treatment group were significantly higher at week 2 than
cholesterol over the course of the study (P 0.78; Table 4).
the control group. Overall, the treatment group showed a
significant decline in heart rate compared to the control High-density lipoprotein cholesterol was similar at baseline
group (P 0.0003). for the two groups (P 0.49) and was log transformed for
analysis.24 Levels did not differ between groups at any time
Most electrocardiograms (ECGs) were considered within
the normal range. Clinically insignificant deviations in ECGs except month 6, when the treatment group had significantly
International Journal of Obesity
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1550
concentrations were not different after 9 months in control
Change in Body Weight (kg)
1 subjects but decreased by nearly 18% (P 0.017) after 9
months in the treatment group. Insulin sensitivity as
0
assessed by the HOMA index was unchanged in the control
Change in weight (kg)
-1 *
group, but improved by 23% (P 0.004) in the group
-2
consuming the dietary supplement with ephedra and
-3 Control
* caffeine. The absolute and proportional (percent) changes
Treatment
-4
of plasma glucose, insulin, and insulin sensitivity in the
-5
treatment group were not related to either the absolute or
-6
the percent change of body fat mass by simple regression
-7 analysis. Baseline plasma leptin concentrations averaged
-8 approximately 20 22 ng/ml in both groups. Leptin was
1.5 mo 3 mo 6 mo 9 mo unchanged in the control group, but decreased by
Control 0.02 -0.98 -1.67 -2.25
25.875.2% (P 0.0004) after 9 months in the supplement-
Treatment -0.68 -3.04 -6.31 -7.18
treated subjects (Table 5). The absolute and proportional
changes of leptin were significantly correlated with the
Change in Fat (kg)
absolute and percent changes of BMI (P 0.0005 and 0.001,
2
respectively). Circulating adiponectin levels at baseline were
1
similar in the control and treatment groups and remained
0 unchanged after 9 months. Initial concentrations of total
Change in fat (kg)
+
plasma ghrelin averaged approximately 700 pg/ml in both
-1
the control and supplement-treated groups, and were un-
++ Control
-2
Treatment changed after 9 months in both groups.
-3
-4
-5 Diet, physical activity, and quality of life
++ +++
At the beginning of the study, reported daily energy intake
-6
1.5 mo 3 mo 6 mo 9 mo for the treatment group was 8079 kJ (1930 kcal) compared to
Control 0.8 1.42 -0.35 -0.99 6614 kJ (1580 kcal) for the control group (P 0.005). After 9
Treatment -0.22 -1.58 -5.34 -5.33
months, the reported mean daily energy intake in the
treatment group decreased to 6367 kJ (1521 kcal) (Po0.006
Figure 2 Changes in body weight and body fat in control and treatment
groups.
compared to baseline value), while the mean daily caloric
intake in the control group was 6455 kJ (1542 kcal). Two
subjects reported energy intake o2512 kJ/day (o600 kcal/
higher HDL than the control group (Po0.0001). The ratio of
day) at both the beginning and end of the study, and six
total cholesterol to HDL cholesterol was significantly lower
additional subjects reported energy intakes of o4186 kJ/day
at months 1.5, 3, 6, and 9 for the treatment group compared
(o1,000 kcal/day).
to the control group, as was the trend over time (P 0.004).
The two groups reported similar levels of physical
Low-density lipoprotein cholesterol was significantly higher
activity at the beginning and end of the study, and no
at baseline in the control group compared to the treatment
differences were found between groups. Six of the seven
group, and both groups showed a decline over the course of
quality of life domains assessed did not differ between
the intervention. Triglyceride values were log transformed
groups. Subjects in the treatment group reported a
for analysis. Values for the treatment group consistently
significantly lower vitality index at baseline compared to
declined over the study, while values in the control group
those in the control group, and the vitality scores rose
increased from baseline to 6 months and then declined at 9
significantly for the treatment group while no change was
months. Differences between groups were significant over
found in the control group.
time (P 0.050).
Among all other clinical chemistry, blood histology and
urinalysis values monitored, no clinically significant changes
Self-reported symptoms
were found.
The treatment group reported significantly more dry mouth,
nervousness, and palpitations than the control group
Endocrine parameters (Table 6). None of the symptoms were severe and no subject
Fasting glucose levels were unchanged in the control group was dropped from the study due to such self-reports. The
over time, while values in the treatment group declined treatment group also reported significantly more increased
significantly (P 0.001) (Table 5). Fasting plasma insulin energy and decreased appetite than the control group.
International Journal of Obesity
Multinutrient supplement with low-dose ephedra and caffeine
RM Hackman et al
1551
Table 3 Blood pressure and heart rate in control and treatment groups
Blood pressure (mm Hg) Control Treatment
7s.e.m. 7s.e.m.
n mean n mean P
Systolic
Baseline 30 105 1.5 29 107 1.7 0.226
Week 1 31 108 1.4 29 108 1.3 0.767
Week 2 30 106 1.5 28 108 1.3 0.252
Week 4 30 108 1.5 28 110 1.5 0.112
Week 6 29 106 3.5 26 107 2.0 0.589
Month 2 26 108 1.4 24 111 2.0 0.113
Month 3 23 107 2.1 22 111 2.1 0.035
Month 4 23 110 1.7 21 111 2.4 0.645
Month 5 23 109 2.7 20 112 2.7 0.350
Month 6 23 110 2.3 19 110 1.6 0.774
Month 7 23 110 1.3 19 112 2.8 0.770
Month 8 23 111 1.9 18 111 2.8 0.814
Month 9 23 112 2.1 17 109 2.2 .075
Overall: Group P 0.609
Group Visit P 0.381
Diastolic Baseline 30 68 1.3 29 70 1.3 0.079
Week 1 31 68 1.3 29 69 1.3 0.309
Week 2 30 68 0.9 28 68 1.1 0.978
Week 4 30 68 1.5 28 68 1.3 0.845
Week 6 29 68 1.5 26 69 1.4 0.322
Month 2 26 68 1.4 24 70 1.4 0.195
Month 3 23 68 1.5 22 70 1.7 0.139
Month 4 23 69 1.3 21 71 1.3 0.375
Month 5 23 68 1.5 20 72 1.8 0.065
Month 6 23 69 1.5 19 69 1.4 0.510
Month 7 23 68 1.3 19 71 1.6 0.074
Month 8 23 70 1.0 18 69 1.7 0.408
Month 9 23 70 1.5 17 71 1.9 0.689
Overall: Group P 0.518
Overall: Group Visit P 0.471
Heart ratea (beats/min) Control Treatment
n Mean Lower 95% CI Upper 95% CI n Mean Lower 95% CI Upper 95% CI P
Baseline 30 68.9 66.7 71.3 29 66.7 64.5 69.0 0.174
Week 1 31 76.9 74.5 79.5 29 72.7 70.3 75.2 0.018
Week 2 30 74.2 71.7 76.7 29 78.6 76.0 81.3 0.016
Week 4 30 75.3
& 2006 Nature Publishing Group All rights reserved 0307-0565/06 $30.00