Content uploaded by Alexander George Panossian
Author content
All content in this area was uploaded by Alexander George Panossian
Content may be subject to copyright.
Clinical trial of Rhodiola rosea L. extract
SHR-5 in the treatment of mild to moderate
depression
V. DARBINYAN, G. ASLANYAN, E. AMROYAN, E. GABRIELYAN, C. MALMSTRO
¨
M,
A. PANOSSIAN
Darbinyan V, Aslanyan G, Amroyan E, Gabrielyan E, Malmstro
¨
m C, Panossian A. Clinical trial
of Rhodiola rosea L. extract SHR-5 in the treatment of mild to moderate depression. Nord J
Psychiatry 2007;61:343348. Oslo. ISSN 0803-9488.
The objective of this study was to assess the efficacy and safety of standardized extract SHR-5 of
rhizomes of Rhodiola rosea L. in patients suffering from a current episode of mild/moderate
depression. The phase III clinical trial was carried out as a randomized double-blind placebo-
controlled study with parallel groups over 6 weeks. Participants, males and females aged 18
70 years, were selected according to DSM-IV diagnostic criteria for depression, the severity of
which was determined by scores gained in Beck Depression Inventory and Hamilton Rating
Scale for Depression (HAMD) questionnaires. Patients with initial HAMD scores between 21
and 31 were randomized into three groups, one of which (group A: 31 patients) received two
tablets daily of SHR-5 (340 mg/day), a second (group B: 29 patients) received two tablets twice
per day of SHR-5 (680 mg/day), and a third (group C: 29 patients) received two placebo tablets
daily. The efficacy of SHR-5 extract with respect to depressive complaints was assessed on days 0
and 42 of the study period from total and specific subgroup HAMD scores. For individuals in
groups A and B, overall depression, together with insomnia, emotional instability and
somatization, but not self-esteem, improved significantly following medication, whilst the
placebo group did not show such improvements. No serious side-effects were reported in any of
the groups AC. It is concluded that the standardized extract SHR-5 shows anti-depressive
potency in patients with mild to moderate depression when administered in dosages of either 340
or 680 mg/day over a 6-week period.
Adaptogens, Rhodiola rosea, Depression, Placebo-controlled trial, Double-blind parallel-group
trial.
A. Panossian, Swedish Herbal Institute Research &Development, Spa˚rva
¨
gen 2, SE- 432 96
A
˚
skloster, Sweden, E-mail: alexander.panossian@shi.se; Accepted 21 July 2006.
D
epressive symptoms are not only disabling but are
also very common. It is estimated that around 10%
of the adult population suffer from such disorders,
although some studies indicate that the true figure may
be higher. Typically, mild to moderate depression pre-
sents itself in the form of mood disturbance, lack of
mental energy, sleep disturbance, low self-esteem and a
wide variety of somatic complaints. In those cases that
do not present serious melancholy or suicidal risk, it
would be preferable to treat the disorder with anti-
depressive medications that do not exhibit any of the
disturbing side-effects associated with the standard
drugs, such as tricyclic anti-depressants, selective ser-
otonin or noradrenaline reuptake inhibitors (13).
Rhodiola rosea L. is a well-known adaptogen that
induces a state of non-specific resistance to damaging
effects of various stressors (46). Previously, it has been
demonstrated that when an extract of R. rosea is
administered together with tricyclic anti-depressants
there is a marked reduction in the side-effects of the
drugs and a positive effect on psychopathological
symptoms in patients with psychogenic depression
(7, 8). However, we have not been able to find any
publications concerning the clinical trial of the anti-
depressive effect of mono-therapy with R. rosea.
The aim of the present study was to investigate the
direct anti-depressive effect of SHR-5, a standardized
extract of R. rosea rhizome, in patients suffering from
mild to moderate depression. A double-blind, placebo-
controlled, randomized phase III pilot study was carried
out in order to determine the therapeutic efficacy and
safety of SHR-5 extract compared with placebo.
# 2007 Taylor & Francis DOI: 10.1080/08039480701643290
The relationship between different doses of the adapto-
gen (340 and 680 mg/day of extract per day) and their
anti-depressant efficacy compared with placebo were
examined.
Material and Methods
Study protocols were submitted to and approved by the
Ethics Committee of the Armenian Drug and Medical
Technology Agency of the Ministry of Health of the
Republic of Armenia. Studies were performed in com-
pliance with the revised declaration of Helsinki (9) and
written informed consent was obtained from all partici-
pants.
Study drugs
All test medications were manufactured according to
Good Manufacturing Practice (GMP) by the Swedish
Herbal Institute (SHI: Gothenburg, Sweden) and pre-
sented in the form of white tablets coated with sugar.
The verum tablets (400 mg) each contained 170 mg of
Rhodiola extract SHR-5. Placebo tablets (400 mg) were
prepared using the inactive ingredients that were iden-
tical to verum tablets except 170 mg of lactose. Verum
and placebo tablets were identical in appearance and
organoleptic properties. Individual medication packs
were provided to subjects in the form of plastic jars
containing 100 tablets and bearing a code number
(randomly encoded in a drug list) on a label that was
designed in accordance with the valid guidelines for
Good Clinical Practice (GCP). In addition to the main
label, each jar had a small supplementary label with
information on dosage regimen (two or four) were to be
taken daily. Duplicate jars, each containing 100 tablets,
were provided to those patients required to take four
tablets daily.
The code was broken after completion of the study
and after statistical analyses had been performed.
Study design
The investigation was a double-blind, randomized,
parallel-group evaluation of SHR-5 extract vs. placebo.
Male and female subjects, aged between 18 and 70 years
and diagnosed with mild or moderate depression ac-
cording to the Diagnostic and Statistical Manual of
Mental Disorders (DSM-IV) criteria (10), attending the
in- or outpatients sections of the Erebouni Medical
Center, Department of Neurology, Armenian State
Medical University, Yerevan, Armenia, were considered
for inclusion in the study. Since eligible individuals were
required to have clinically significant depression, only
patients with initial scores of ]13 on the Beck Depres-
sion Inventory (BDI: 13-item short form) and of ]21
on the Hamilton Rating Scale for Depression (HAMD;
21-item version) (1113) were selected. The BDI and
HAMD tests were performed by skilled physicians using
Armenian and Russian language versions of the ques-
tionnaires as appropriate in order to avoid possible
misinterpretations.
Subjects were excluded from the study if: 1) there was
evidence of a previously documented or reported
attempt to commit suicide, 2) if they scored ]1on
BDI item H or ]2 on HAMD item 3 (suicidal
tendency), 3) if they presented a total HAMD score
above 31, 4) if they presented progressive organic or
metabolic brain syndrome, or compulsive, schizophrenic
or other delusive disorders, or 5) if they were pregnant or
lactating. Patients suffering from, and/or receiving
medication for serious chronic illnesses (including car-
diovascular diseases and diabetes) were also excluded
because of the possibility of drug interactions.
Prior to the commencement of the study, all included
subjects underwent routine blood tests and a general
medical examination. This was followed by a 2-week
run-in period, during which time no medication was
provided for any patients, including those who were
receiving, or who had recently received, anti-depressant
or psychotropic drugs. After this time, patients were
randomly distributed amongst one of three groups,
group A receiving two tablets once daily of SHR-5
standardized extract (340 mg/day), group B receiving
two tablets twice daily of SHR-5 standardized extract
(680 mg/day), and group C receiving two placebo tablets
once daily. The distribution of subjects within groups
was carried out, according to the principles of total
randomization, whereby each patient was randomly
assigned an integer 190.
The number selected provided the identification code
for the patient and the drug code number (randomly
encoded in a drug list), and both were recorded in a
protocol and in the journal for each patient in order to
permit subsequent identification. Information concern-
ing the content of extracts became available to the
investigators and volunteers only after completion of the
study and final statistical analysis of the results.
A sample size of 30 subjects per group was estimated
to be sufficient for a 15% significance level by assuming
a 30% effect difference between the treatment groups
and the placebo group over the study period, with a 95%
confidence interval. A total of 91 patients, 31 in group A
and 30 each in groups B and C, formed the initial
baseline set.
BDI and HAMD questionnaires were administered to
subjects at the start (day 0) and the end (day 42) of the 6-
week period of treatment. Throughout the study phase,
patients were followed-up by telephone contact or
personally by attending physicians, and any adverse
events were recorded. At the end of the study, patients
underwent further routine blood tests and a general
medical examination. All complaints arising from the use
of the study medications were documented by the
V. D
ARBINYAN ET AL
.
344 NORD J PSYCHIATRY ×VOL 61 ×NO 5 ×2007
attending principal investigator, physicians and monitors
in order to assess the safety of the materials employed.
Compliance with the regime of medication by individuals
was determined by collecting the used jars and unused
tablets on the last visit, and by questioning of the patients
by the investigator. Deviations from required protocols
were recorded together with the reason and the date.
All data, including signed written consents, personal
information, clinical reports, medical histories, results of
BDI and HAMD assessments, adverse reaction reports
and reasons for study termination, were collected in a
journal for each patient, which was signed by the
principal investigator and the study monitor.
Parameters for measuring efficacy of treatment
The efficacy of medication with respect to depressive
symptoms was assessed from the symptom scores in the
BDI and HAMD questionnaires administered at the
start and at the end of the treatment period. The primary
efficacy variable was represented by the change in total
BDI and HAMD scores over the study period. Second-
ary efficacy variables were classified as the changes, over
the study period, in the HAMD symptom indicator
subgroups I (points 4, 5 and 6: insomnia), II (points 9
and 10: emotional instability), III (points 1216: soma-
tization and hypochondria) and IV (points 17 and 18:
self-esteem), which reflect the different somatic and
physiological aspects of mild and moderate depression.
Statistical methods
The data for each subject, identified only by the entry
number of the patient, were entered into the study
database. Data management and statistical analyses
were performed using PRISM Statistical Software (ver-
sion 2.01, 1996). The statistical analyses were performed
according to Student’s t-test and the Wilcoxon non-
parametric two-tailed rank test. Pearson’s correlation was
used to test for correlation between variables. The study
was planned with adaptive interim analyses (14) following
treatments involving three groups of 30 patients. The
overall type 1 error rate was set at a 0.05, implying a
nominal level of a 0.0299 for the assessment of the
statistical significance in the interim analyses.
It was estimated that a sample size of 30 patients per
group would permit detection of a standardized treat-
ment difference of 0.50 with a power of 80% (two-sample
t-test: a 0.05: one-sided).
Results
A total of 91 patients, with mild to moderate depression,
were initially selected according to the inclusion and
exclusion criteria, and 89 were randomized to one of
three treatment groups, treatment group A and B and a
placebo group C. Two patients dropped out of the study
for non-medical reasons. No patients reported serious
adverse side-effects of the medications and none were
excluded through non-compliance with the prescribed
treatment. The distribution by sex and the mean ages of
the 89 participants who completed the study are
presented in Table 1. No statistically significant differ-
ences in demographic data could be determined between
the groups.
Tables 2 and 3 show the HAMD scores at the start of
the study (day 0) and following treatment (day 42),
which involved the administration of tablets containing
either SHR-5 standardized extract (340 mg/day*group
A; 680 mg/day*group B) or placebo (group C). There
were no statistically significant differences (n.s.) in the
mean HAMD and BDI scores between the three groups
prior to the start of the treatment. However, the scores
obtained by subjects in groups A and B after 42 days of
treatment with SHR-5 indicated the presence of sig-
nificantly (PB0.0001) reduced total levels of symptoms
(Fig. 1). Thus, in groups A and B, the mean total
HAMD scores declined from 24.52 to 15.97 (PB0.0001),
and from 23.79 to 16.72 (PB0.0001), respectively, whilst
the score of the placebo group C showed no such
improvement (mean HAMD score of 24.17 before
treatment vs. 23.41 after treatment (P0.3306). The
mean total HAMD scores for groups A and B at the end
of the study were statistically significantly different (PB
0.001) from that of group C (Tables 2 and 3).
Fig. 2 shows the efficacy of the SHR-5 standardized
extract, as administered to subjects suffering from mild
to moderate depression, with respect to changes in
insomnia, emotional instability and the level of somati-
zation (Fig. 2ac, respectively), whilst the intensity of
these symptoms of depression, as determined from the
scores in the HAMD symptom indicator subgroups,
were similar in the treatment groups A and B and the
placebo group C prior to the start of treatment.
Statistical significant improvements were shown in the
subjects of both treatment groups, but not of the placebo
group.
In contrast, the initially low levels of self-esteem
determined in subjects of all groups at the start of the
study (Fig. 2d), remained unchanged after 42 days of treat-
ment with two tablets daily of SHR-5 (group A) or with
placebo (group C), but improved significantly (2.690
Table 1. Characteristics of selected patients suffering from
mild and moderate depression randomized into treatment
groups within the study.
Group Males Females
Age (years),
mean9s
Group A (verum) (n 31) 10 (32.3%) 21 (67.7%) 44.9911.5
Group B (verum) (n29) 12 (41.4%) 17 (58.6%) 44.60925.49
Group C (placebo) (n29) 14 (48.3%) 15 (51.7%) 42.80912.87
s, standard deviation.
R
HODIOLA AND THE TREATMENT OF DEPRESSION
NORD J PSYCHIATRY ×VOL 61 ×NO 5×2007 345
before treatment and 1.897 after treatment, P 0.0002)
after medication with four tablets daily of SHR-5 (group B).
Discussion
Previous research concerning the pharmacological ef-
fects of R. rosea extract has focused mainly on its ability
to increase mental performance under stress and to
relieve temporary fatigue. During the last decade or so,
R. rosea has been used extensively as a herbal remedy for
negative stress reactions, and the number of physicians
prescribing the drug, especially in northern
Europe, has increased. Our own clinical experience,
combined with experience from colleagues using
Table 2. Total scores obtained by patients suffering from mild and moderate depression in treatment groups A and B (verum) and
the placebo group C submitted to Hamilton Rating Scale for Depression (HAMD) questionnaires before (day 0) and after (day
42) treatment.
Group A (verum),
two tablets SHR-5/day
Group B (verum),
four tablets SHR-5/day Group C (placebo)
Day of treatment period 0 42 0 42 0 42
Number of patients questioned 31 31 29 29 29 29
Mean HAMD score 24.52 15.97 23.79 16.72 24.17 23.41
s 2.249 4.637 1.698 4.174 1.692 3.803
SE 0.4039 0.8328 0.3154 0.7751 0.3142 0.7063
95% confidence interval
Lower limit 23.69 14.27 23.15 15.14 23.53 21.97
Upper limit 25.34 17.67 24.44 24.44 24.82 24.86
Coefficient of variation 9.17% 29.04% 7.14% 24.96% 7.00% 16.24%
Within group comparison of mean HAMD scores
before (day 0) and after (day 42) treatment
Comparison within group A B C
P-value (paired t-test) B0.0001 B0.0001 0.33
Difference between mean scores Highly significant Highly significant Not significant
Therapeutic effect demonstrated Yes Yes No
Between group comparison of mean HAMD
scores before treatment (day 0)
Comparison between groups A and B A and C B and C
P-value (paired t-test) 0.077 0.51 0.19
Difference between mean scores Not significant Not significant Not significant
Between group comparison of mean HAMD
scores after treatment (day 42)
Comparison between groups A and B A and C B and C
P-value (paired t-test) 0.34 PB0.0001 PB0.0001
Difference between mean scores Not significant Highly significant Highly significant
s, standard deviation.
042
0
10
20
30
ns
***
***
Group C (placebo)
Group A (verum) two tablets SHR-5/day
Group B (verum) four tablets SHR-5/day
Groups of patients
erocsDMAH
042
0
5
10
15
ns
***
***
Groups of patients
erocsIDB
Fig. 1. The effect of treatment with SHR-5 (groups A and B) vs. placebo (group C) as measured by the primary efficacy variable,
namely, the change in total Hamilton Rating Scale for Depression (HAMD) and Beck Depression Inventory (BDI) scores
between the start (day 0) and the end (day 42) of medication. (For the within-group comparisons of mean scores at days 0 and 42
the significance of the differences are shown as: ns, not significant, and ***, significantly different with PB0.0001).
V. D
ARBINYAN ET AL
.
346 NORD J PSYCHIATRY ×VOL 61 ×NO 5 ×2007
Table 3. Total scores obtained by patients suffering from mild and moderate depression in treatment groups A and B (verum) and
the placebo group C submitted to the Beck Depression Inventory (BDI) questionnaire prior to (day 0) and after (day 42) treatment.
Group A (verum),
two tablets SHR-5/day
Group B (verum), four
tablets SHR-5/day Group C (placebo)
Day of treatment period 0 42 0 42 0 42
Number of patients questioned 31 31 29 29 29 29
Mean HAMD score 12.23 7.097 10.38 4.750 11.33 9.897
s 5.214 3.636 4.296 2.744 2.440 5.659
SE 0.9364 0.6531 0.7978 0.5185 0.4455 1.051
95% confidence interval
Lower limit 10.31 5.763 8.745 3.686 10.42 7.744
Upper limit 14.14 8.431 12.01 5.814 12.24 12.05
Coefficient of variation 42.64% 51.24% 41.39% 57.76% 21.53% 57.18%
Within group comparison of mean HAMD scores
before (day 0) and after (day 42) treatment
Comparison within group A B C
P-value (paired t-test) PB0.0001 PB0.0001 0.21
Difference between mean scores Highly significant Highly significant Not significant
Therapeutic effect demonstrated Yes Yes No
Between group comparison of mean
HAMD scores before treatment (day 0)
Comparison between groups A and B A and C B and C
P-value (paired t-test) 0.14 0.40 0.72
Difference between mean scores Not significant Not significant Not significant
Between group comparison of mean
HAMD scores after treatment (day 42)
Comparison between groups A and B A and C B and C
P-value (paired t-test) 0.0075 PB0.0001 PB0.0001
Difference between mean scores Significant Highly significant Highly significant
s, standard deviation.
insomnia
day 0 day 42
day 0 day 42
day 0 day 42
day 0
day 42
0.0
2.5
5.0
7.5
***
ns
***
Groups of patients
puorgbus-erocsDMAH
I
emotional instability
0
1
2
3
4
***
ns
***
Groups of patients
p
uorgbus-e
r
o
c
sDMA
H
II
somatization
0
5
10
15
***
ns
**
Groups of patients
HAMD score - subgroup
III
self-esteem
0
1
2
3
ns
ns
***
Group A (verum) two tablets SHR-5/day
Group C (placebo)
Group B (verum) four tablets SHR-5/day
Groups of patients
pu
o
rgbus-e
ro
c
s
DMAH
VI
Fig. 2. The effect of treatment with SHR-5 (groups A and B) vs. placebo (group C) as measured by the secondary efficacy
variables, namely, the change in the four Hamilton Rating Scale for Depression (HAMD) subgroups: (a) insomnia, (b) emotional
instability, (c) somatization and (d) self-esteem (corresponding to the different somatic and physiological aspects of mild and
moderate depression) between the start (day 0) and the end (day 42) of medication. (For the within-group comparisons of mean
scores at days 0 and 42 the significance of the differences are shown as: ns, not significant, **, significantly different at P B0.001,
and ***, significantly different with PB0.0001).
R
HODIOLA AND THE TREATMENT OF DEPRESSION
NORD J PSYCHIATRY ×VOL 61 ×NO 5×2007 347
R. rosea as a treatment, is that the positive effect of the
drug is mediated by mood stabilization and energy
restoration. This is the first randomized controlled study
that confirms the potential anti-depressive effects of
R. rosea as mono-therapy.
The present clinical study has shown that the stan-
dardized extract SHR-5 from R. rosea possesses a clear
and significant anti-depressive activity in patients suffer-
ing from mild to moderate depression. When adminis-
tered in a dosage of two tablets, each containing 170 mg
of extract, daily over a 6-week period, statistical
significant reduction in the overall symptom level of
depression as well as in specific symptoms of depression,
such as insomnia, emotional instability and somatiza-
tion, could be demonstrated. In higher doses, four
tablets per day over a 6-week period, an additional
positive effect could be shown, the level of self-esteem
increased significantly. No side-effects resulting from
treatment with SHR-5 could be detected in any group of
the groups.
Future studies with follow up period up to 12 weeks as
well larger multicenter studies, could eventually answer
the question of how SHR-5 compares with the classic
anti-depressants in efficacy.
Such studies could also focus on of the pharmacolo-
gical mechanisms for the mode of action for R. rosea in
mild to moderate depression.
Acknowledgements*This study was funded with project grants from
the Swedish Herbal Institute. The funding sponsor, however, had no
role in any practical aspect of the study including collection,
management, analysis and interpretation of the data, or review and
approval of the manuscript. The study was conceived by AP, the
protocol was formulated by VD, and the ethical application was
administered by GA. VD enrolled patients and performed the study
interventions, the assays and, together with EA, the data analysis. CM
drafted the manuscript, and all authors were involved in its critical
appraisal and final approval. The authors wish to thank the volunteers
for their invaluable help with the study. (The use of SHR-5 for the
treatment of depression is the subject of pending patent application by
Swedish Herbal Institute.)
References
1. Wells KB, Golding JM, Burnham MA. Psychiatric disorders in a
sample of the general population with and without chronic medical
conditions. Am J Psychiatry 1988;
/145:/97681.
2. Von Korff M, Shapiro S, Burke JD, Teitlebaum M, Skinner EA,
German P, et al. Anxiety and depression in a primary care clinic:
Comparison of diagnostic interview schedule, general health
questionnaire, and practitioner assessments. Arch Gen Psychiatry
1987;
/44:/1526.
3. Shear MK, Schulberg HC. Anxiety disorders in primary care. Bull
Menn Clin 1995;
/59:/A7385.
4. Saratikov AS, Kranov EA, Chnikina LA, Duvidson LM, Sotova
MI, Marina TF, et al. Rhodiolosid, ein neues Glykosid aus
Rhodiola rosea und seine Pharmakologischen Eigenschaften. Die
Pharmazie 1968;
/23:/3925.
5. Saratikov AS. Golden root (Rhodiola rosea). Tomsk: Tomsk
University Press; 1973 (Russ.), 1974 (Engl.).
6. Saratikov AS, Krasnov EA. Rhodiola rosea (Golden root). Tomsk:
Tomsk University Publishing House; 2004.
7. Brichenko VS, Kupriyanova IE, Skorokhodova TF. The use of
herbal adaptogens together with tricyclic anti-depressants in
patients with psychogenic depressions. In: Goldsberg ED, editor.
Modern problems of pharmacology and search for new medicines,
Volume 2. Tomsk: Tomsk State University Press; 1986. p. 5860.
8. Brichenko VS, Skorokhodova TF. Herbal adaptogens in rehabili-
tation of patients with depressions. In: Clinical and organizational
aspects of early manifestations of nervous and mental diseases.
Barnaul; 1987. p. 15.
9. World Medical Association Declaration of Helsinki. Ethical
principles for medical research involving human subjects. 18th
WMA General Assembly, Helsinki (Finland), June 1964 and 52nd
WMA General Assembly, Edinburgh (Scotland) October 2000.
10. American Psychiatric Association. Diagnostic and statistical
manual of mental disorders. Washington D.C.: APA; 1994.
11. Schwab J, Bialow M, Clemmons R, Martin P, Holzer C. The Beck
Depression Inventory in medical inpatients. Acta Psychiatr Scan
1967;
/43:/25566.
12. Schwab JJ, Bialow MR, Clemmons RS, Holzer CE. Hamilton
Rating Scale for Depression with medical in-patients. Br J
Psychiatry 1967;
/113:/838.
13. Hamilton M. A rating scale for depression. J Neurol Neurosurg
Psychiatry 1960;
/23:/626.
14. Bauer P, Kohne K. Evaluation of experiments with adaptive
interim analyses. Biometrics 1994;
/50:/102941.
V. Darbinyan, Department of Neurology, Armenian State Medical
University, Yerevan, Armenia.
G. Aslanyan, Scientific Centre of Drug and Medical Technology
Expertise, 15, Moscovyan Street, Yerevan, 375001, Armenia.
E. Amroyan, Scientific Centre of Drug and Medical Technology
Expertise, 15, Moscovyan Street, Yerevan, 375001, Armenia.
E. Gabrielyan, Scientific Centre of Drug and Medical Technology
Expertise, 15, Moscovyan Street, Yerevan, 375001, Armenia.
C. Malmstro
¨
m, The PBM Clinic, Institute of Health Competence,
Arenava
¨
gen 41, 12177 Stockholm, Sweden.
A. Panossian, Swedish Herbal Institute Research &Development,
Spa˚rva
¨
gen 2, SE- 432 96 A
˚
skloster, Sweden.
V. D
ARBINYAN ET AL
.
348 NORD J PSYCHIATRY ×VOL 61 ×NO 5×2007