sign in sign up
<<
Home , Anxiolytic, Withania somnifera

Indian Journal of Psychiatry, 2000, 42 (3), 295-301

A DOUBLE-BLIND, PLACEBO-CONTROLLED EVALUATION OF THE ANXIOLYTIC EFFICACY OF AN ETHANOLIC EXTRACT OF SOMNIFERA

CHITTARANJAN ANDRADE, ANITHA ASWATH, S.K. CHATURVEDI, M. SRINIVASA & R. RAGURAM

ABSTRACT

A double-blind, placebo-controlled study was conducted to evaluate the efficacy an ethanolic extract of Aswagandha (Withania somnifera), in patients with ICD-10 disorders. The sample comprised 39 subjects, of whom 20 received the and 19 received placebo. The two groups were sociodemographically and clinically similar at baseline. At 2 and 6 weeks follow-up, data from approximately 85% of patients in each group were available for analysis. Statistical trends favouring the drug were observed at both time points. At 6 weeks, significantly more patients met a priori response criteria in the drug group (88.2%) as compared with the placebo group (50%). The drug was well-tolerated and did not occasion more adverse effects than did placebo. It is concluded that this ethanolic extract of Withania somnifera has useful anxiolytic potential and merits further investigation. Key words: , generalized anxiety disorder, adjustment disorder with anxiety, , Aswagandha, Ayurvedic treatment, herbal therapy, Anxiety disorders are common, and have somnifera) has been used for centuries in a lifetime prevalence that exceeds 15% in the Ayurvedic medical practice to reduce symptoms general population (Kessler et al., 1994). of anxiety and stress (Iyengar, 1981; Handa,1995; Conventional anxiolytic treatments include the Satyavati,1995). In laboratories in the USA, , in low Withania somnifera extracts have been shown doses, and . All these three categories to possess high affinity for GABA receptors (Cott of drugs are associated with disadvantages in et al.,1994). The experimental drug is an the management of anxiety. Benzodiazepines ethanolic extract of Withania somnifera. In Indian and antidepressant drugs occasion sedation and research, this extract has been shown to have impaired psychomotor performance; while GABA-mimetic properties (Mehta et al .1991; therapy results in drug Kulkarni et al.,1993), which might conceivably dependence, antidepressant therapy is mediate its demonstrated associated with anticholinergic and/or other (Kulkarni et al.,1993; Kulkarni & George,1996) adverse effects. Antidepressant drugs and and antikindling (Kulkarni et al., 1994) actions. buspirone may both produce initial worsening of Since GABA agonism has b< en linked to anxiety (Andrade,2000). There is therefore a anxiolysis (Stahl,1998), these findings support need for the development of anxiolytic therapies the recommendation in A"urveda that that carry fewer disadvantages to the patient. Aswagandha be used for tranquillization. In , Aswagandha (Withania Extracts of Aswagandha may therefore be useful

295 CHITTARANJAN ANDRADE et al. in anxiety disorders that present in psychiatric recruited patients. The project protocol received practice. the approval of the Ethics Committee in the In an unpublished open clinical trial (data institution of the first author. on file; available upon request from Patients were randomized to receive manufacturer), Dave at the Rasiklal Shah either ethnolic extract of Aswagandha (250 mg/ Sarvajanik Hospital, Gujarat, studied the efficacy tablet) or placebo, administered as identical and safety of ethnolic extract of Aswagandha tablets in the dose of 2 tablets twice a day. The 250 mg twice daily in 50 patients with anxiety first follow-up was conducted after 2 weeks, at disorders. By the end of the first month of which time a decision was taken to lower or raise treatment, 36 (72%) of patients showed the dose based upon the clinical response and moderate to excellent improvement, in about half the adverse effects reported. Subsequently, dose of these cases, benefits were observed within titrations to individual requirements were allowed the first fortnight. These 36 patients were at weekly intervals, subject to a minimum of 2 continued on the drug for a maximum of 18 and a maximum of 10 tablets per day. For ethical months, during which period the was reasons, existing (if any) were withdrawn under psychotherapeutic support. It continued unchanged throughout the course of was observed that the drug satisfactorily the study. No other medication, or any other form attenuated anxiety symptoms and did not of therapeutic intervention, was prescribed occasion any adverse effects of note. during the study. In view of the encouraging laboratory and Patients were assessed at baseline, at clinical results, a double-blind, placebo- week 2, and at week 6 (the treatment endpoint) controlled, dose-ranging study was designed to using the following instruments : assess the anxiolytic efficacy and adverse effects 1. Hamilton Anxiety Scale (Bech et at., 1986); of this ethanolic extract of Aswagandha. assessed by the rater only. 2. Global Rating Scale; assessed separately by MATERIAL AND METHOD patient and rater. This instrument had rating points anchored as follows : 0=no symptoms; The sample comprised consecutive 1=mild symptoms; 2=moderate symptoms; patients diagnosed with ICD-10 generalized 3=severe symptoms; 4=very severe symptoms. anxiety disorder, mixed anxiety and depression, Decimal ratings were encouraged. panic disorder, and adjustment disorder with 3. Systematic assessment for treatment anxiety. The sample was recruited from two emergent effects (SAFTEE) symptom checklist clinical units in a large mental hospital setup, (Levine & Schooler, 1986), assessed by the rater and from a private psychiatric hospital, both only, and at weeks 2 and 6 only. located in urban India. The study was conducted Patients and raters were alike blind to the on an outpatient basis. treatment allocation until after the study Patients were recruited only if their treating endpoint. consultant considered them to be sufficiently Statistical analysis : Baseline data were analyzed symptomatic to warrant further pharmaco- for all recruited patients. Efficacy and adverse therapeutic intervention Further selection criteria effects data were analyzed only for those patients included an absence of significant past or current who provided follow-up assessments at week 2 history of cigarette smoking, drug or and beyond. The last observation carried forward abuse, or medical or psychiatric illness that may (LOCF) method was employed for all patients have prejudiced the diagnosis or ratings of who completed the first-fortnight assessment but anxiety, or the response to anxiolytic medication. who dropped out subsequently. Response was Informed consent was obtained in writing from defrned a priori as a reduction in Hamilton

296 ANXIOLYTIC EFFICACY OF ETHANOLIC EXTRACT OF WITHANIA SOMNIFERA

Anxiety score to 12 or below, associated with a TABLE 2 global rating from both patient and rater of not DIAGNOSTIC BREAKUP OF DRUG (N-20) AND PLACEBO (N=19) GROUPS more than 1 (mild symptoms). Proportions were compared between Diagnosis Drug group Placebo groups using the X2 test (with Yates continuity Generalized 13 11 anxiety disorder correction for 2x2 contigency tables), or the Mixed anxiety 3 5 Fisher's exact 2-tail probability test when the and depresssion requirements for X2 testing were not met. Means Adjustment 2 1 were compared between groups using the disorder with anxiety independent sample student's 't' test, with Panic disorder 2 2 modified degrees of freedom wherever indicated to correct for heterogeneity of variances as TABLE 3 BASELINE CLINICAL DESCRIPTION OF DRUG determined by the Fmax test. When distributions (N=20) AND PLACEBO (N=19) GROUPS* were non-normaJ, mean ranks between groups were compared using the Mann-Whitney U test Variable Drug group Placebo Significance* with z corrected for ties. Means were compared Duration of 0.5-40 0.2-20 z=0.21,NS illness (years) 7.4(10.3) 6.4(6 3) between groups and across time using repeated Hamilton 15-28 15-28 t?0.72. d.f.=37 measures multivariate analysis of variance Anxiety score 17.3(6.4) 18.6(4.5) NS (RMANOVA); Pillai's trace was selected as the Global 1-4 2-4 t=1.10, test criterion. rating 2.6 (0.8) 2.8(0.5) d.f.=31.6,NS (patient) Global 1-4 1-4 t=0.98, RESULTS rating 2.5(0.9) 2.8 (D.6) d f.=37, NS (rater) The sample comprised 39 subjects, of * data presented are range, mean and standard deviation. whom 20 received the experimental drug and # Mann-Whitney U test with z corrected for ties, and independent sample Student's t' test with modified d.f.. whereve 19 received placebo. The sample was 61.5% indicated. to correct for heterogeneity of variances. male, and the mean (standard deviation) [M be assessed on any measure at week 2. At week (SD)] age of the sample was 41.3 (13.8) years. 6, a further 6 experimental drug patients and 7 The sociodemographic and clinical variables of placebo patients withdrew from the study. the sample are presented in tables 1-3. There Reasons for withdrawal are presented in table were no significant differences between 4. There were no significant differences between experimental drug and placebo groups at baseline. groups in individual reasons for withdrawal. At the end of week 2, when the first follow- At baseline, according to the protocol, all up assessments were due, 3 patients belonging to each group dropped out of the study, and TABLE 4 REASONS FOR DROUP OUT IN DRUG (N-20) attempts to trace them to ascertain the reasons AND PLACEBO (N=19) GROUPS for drop put failed. These 6 patients could not Reason for Drug group Placebo TABLE 1 droup out AGE AND SEX DISTRIBUTION IN DRUG (N=20) AND PLACEBO (N=19) GROUPS* Lack of benefit 3 3 Adverse effects 2 3 Variable Drug group Placebo Other" 1 1 Age (in yrs.) 18-68 18-69 Unknown" 3 3 Range * These patients developed a crisis in their personal (MeanS.D.) 41.9(14.4) 40.7(13.4) environments, which necessitated more intensive t=0.27,d.f.= 37, NS medication. Sex: Male 13 11 " These patients dropped out before the first follow-up; all Female 7 8 other patients completed at feaston e follow-up. X,=0.02, d.f.=1, NS (with yates correction)

297 CHITTARANJAN ANDRADE et a/. patients were advised to take two tablets twice significant (Fisher's exact two-tailed daily. At week 2, based on initial response, probability=0.026). patients were advised 2-6 tablets/day with a A quantitative examination of anxiety M(SD) of 4.5 (1.1) tablets in the drug group, and scores across time (Table 5) showed that both 2-6 tablets/day with a M(SD) of 4.3 (1.5) tablets groups improved significantly (Pillai's trace=0.68, in the placebo group. The difference between F=32.13, d.f.=2,30, p<0.001) from week 2 itself. groups was not significant (t=0.57, d.f.=31, NS). There was a main effect for drug that approached At week 6, the patients were receiving 2-9 significance (F=3.35, d.f. = 1,31, p=0.077). tablets/day with a M(SD) of 5.0 (2.1) tablets in Independent comparisons at week 2 (t=1.70, the drug group, and 2-8 tablets/day with a M(SD) d.f.=31, p=0.098) and week 6 (t=1.94, d.f.=31, of 4.7 (2.0) tablets in the placebo group. The p=0.062) revealed statistical trends that favoured difference between groups was again not the experimental drug. significant (t=0.40, d.f.=21, NS). A quantitative examination of changes in Efficacy and data were global scores across time (Table 6) showed available for 17 of the 20 experimental drug significant improvements for both patient (Pillai's patients (85%) and for 16 of the 19 placebo trace=0.73, F=40.56, d.f.=2,30, p<0.001) and patients (84.2%). At the second week, 12 of 17 rater-assessed (Pillai's trace=0.66, F=29.60, experimental drug patients (70.6%) met criteria d.f.=2,30, p<0.001) functioning from week 2 for response, while 6 of 16 placebo patients itself. No significant advantage for experimental (37.5%) were similarly classified; the difference drug was however obtained in either the former was however not statistically significant (X2=2.43, (Pillai's trace=0.02, F=0.25, d.f.=2,30, NS)orthe d.f.=1, NS). At the sixth week, 15 of 17 drug latter (Pillai's trace=0.02, F=0.31, d.f.=2,30, NS) patients (88.2%) met criteria for response, while assessments. Independent testing at each of 8 of 16 placebo patients (50%) were similarly weeks 2 and 6 again did not demonstrate an classified; the difference was statistically advantage for experimental drug. A quantitative analysis of adverse effects TABLE 5 RANGE, M(SD) CHANGE IN HAMILTON ANXIETY (Table 7) showed that 15 experimental drug SCORES ACROSS TIME IN DRUG (N=17) treated patients reported a M(SD) of 1.2 (1.2) AND PLACEBO (N=16) GROUPS" adverse effects at week 2, and 8 patients Week 2 Week 6 reported a M(SD) of 0.25 (0.5) adverse effects Drug 5-22 0-22 at week 6. Corresponding data for placebo- 11.0(4.7) 8.5(5.3) treated patients were 0.9 (1.1) from 15 patients Placebo 4-25 4-25 at week 2, and 0.67 (1.1) from 9 patients at week 14.1(5 6) 12.2(5.6) 6. The difference between groups was not * baseline data have been presented In Table 3. # statistical trends favoured the drug group at both week significant either at week 2 (Mann-Whitney; 2 and week 6. z=0.80, NS) or at week 6 (Mann-Whitney,

TABLE 6 RANGE, M(SD) GLOBAL ASSESSMENTS OBTAINED FROM PATIENTS AND RATER AT WEEK 2 AND 6 IN DRUG (N«17) AND PLACEBO (N*16) GROUPS ## Patients' assessment Rater's assessment Week 2 Week 6 Week 2 Week 6 Drug 0.5-3 0-3 0.5-3 0-3 1.9 (0.8) 1.4 (0.9) 1.8 (0.7) 1.4 (0.9) Placebo 0.5-3.5 0-3.5 0.5-3 0.5-3 2.2 (1.0) 1.8 (1.0) 2.0 (0.9) 1.8 (0.8) * baseline data have been presented in Table 3. # there were no differences between drug and placebo groups at either week 2 or week 6.

298 ANXIOLYTIC EFFICACY OF ETHANOLIC EXTRACT OF WITHANIA SOMNIFERA

TABLE 7 most adverse effects being mild in severity, FREQUENCY OF EXPERIENCE OF ADVERSE appearing early during therapy, and manageable EFFECTS AT WEEKS 2 AND 6 by adjustment in dose; however, there were 5 Number of advers effects reported dropouts due to adverse effects : two in the none 1 2 3 Aswagandha group (reasons : increased gastric At week 2 acidity; withdrawal fatigue), and three in the Drug group 6 3 3 3 Placebo group 6 2 4 1 placebo group (reasons: slowness and dullness; At week 6 giddiness, weakness and fatigue; burning of Drug group 6 2 0 0 eyes, burning while passing urine, and worsening Placebo group 6 111 of obsessive-compulsive symptoms). Z=0.60, NS). Most patients experienced the adverse A qualitative analysis of adverse effects effects early during the study (Table 7). At the is presented in table 8. With the possible study endpoint, the drugs were withdrawn exception of greater drowsiness, heaviness of abruptly. While no formal assessments of head and related symptoms with Aswagandha withdrawal were conducted, clinical records of (during the initial few days only), there were no these patients documented no withdrawal differences between groups in the nature and phenomena in either group. frequency of individual adverse effects. The treatments were in general well-tolerated, with DISCUSSION TABLE 8 ADVERSE EFFECTS REPORTED BY DRUG (N*17) India has a rich tradition in herbal AND PLACEBO (N»16) medicine, and it is unfortunate that little work is Adevrse effect Drug group Placebo group being conducted by allopathic neuroscientists to identify elements in the herbal pharmacopoeia Giddiness 0 2 Drowsiness 2 0 that might be of use in psychiatric disorders. The Dullness 0 1 western pharmaceutical industry is already Heaviness of head 4 1 involved in the research of medicinal plants, and Increased frequency it would be doubly unfortunate if western of migraine 1 0 Fatigue 1 1 scientists patent molecules of psychotropic Withdrawal fatigue 1 0 value, that have been derived from Indian herbs. Chest pain 0 1 This subject has been discussed in greater detail Chest 0 1 elsewhere (Andrade et al.,1998). The present Itching 1 0 Burning eyes 0 1 study was therefore a trail blazer in the field of Burning urine 0 1 anxiety disorders. Numb hands 1 0 Essentially, this study demonstrated a Multiple somatic complaints 0 1 trend for the anxiolytic superiority of drug over Heat in body 0 1 placebo at week 2, and a statistically significant Bitter taste 1 0 superiority at week 6. The advantage for Decreased appetite 1 0 Aswagandha was observed only with Hamilton 1 Increased appetite 1 Anxiety ratings, and with responder analyses, Decreased 2 0 Gastritis 2 1 but not with global ratings; this may have bean Heaviness in stomach 1 1 because the latter had a narrower rating range, Diarrhoea 0 1 and were therefore less sensitive to clinical Menstrual irregularities 0 2 Worsening of obsessive- change. Aswagandha was well tolerated, with compulsive disorder 0 1 adverse effects being comparable to those * Data presented are number of patients reporteing each observed with placebo. At 6 weeks, abrupt adverse effect. withdrawal did not appear to be associated with

299 CHITTARANJAN ANDRADE et al anywithdrawal phenomena. Thus, Aswagandha are warranted, with particular reference to appeared to have several of the advantages but ethanolic extracts. Comparisons with convetional none of the disadvantages of conventional anxiolytic drugs are necessary. The relevance anxiolytic drugs such as the benzodiazepines, of the GABAergic properties of the extract to the tricyclic , and buspirone. This anxioiysis also require investigation. Long-term study therefore confirmed the preliminary studies with structured withdrawal programmes observations of the unpublished open study of should be conducted to document the long-term the experimental drug in patients with anxiety efficacy, safety, and dependence vs withdrawal disorders (Manufacturer, data on file). profile of the treatment. Finally, attemts need to Benzodiazepines produce anxiolysis be made to identify the specific therapeutic within days; antidepressant drugs and buspirone element contained in the extract. reduce anxiety in approximately a fortnight. Speed of response was regrettably not addressed ACKNOWLEDGEMENT in this study. However, it is likely that significant response will be obtained in most patients within The ethanolic extract of Ashwagandha the first fortnight of therepy itself, as Aswagandha (Withania Somnifera), used in this study as the was near-significantJy superior to placebo at this experimental drug is marked in the name of Aswal time -point. Future studies need to address the by the manufacturers. This study was supported magnitude of anxiolysis during the first week to by a research grant from Gufic Ltd., Bombay, the ascertain the position of Aswagandha relative manufacturer of the formulation. to conventional anxiolytic drugs. What might be the mechanism of action REFERENCES of Aswagandha? As reviewed earlier, Aswagandha has been shown to have GABA- Andrade.C. (2000) . mimetic properties (Mehta et al.,1991; Kulkami In : Handbook of Psychiatry, (Ed.) Bhugra.D., et al.,1993); since GABA agonism has been New Delhi: Oxford University Press (in press). linked to anxiolysis (Stah1,1998), it is conceivable that Aswagandha contains a constituent that Andrade.C, Gowda,S. & Chaturvedi, alleviates anxiety by modulating GABA S.K. (1998) Treatment of age-related cognitive neurotransmission. decline with a herbal formulation: a double-blind The high placebo response rate (50%) was study. Indian. Journal of Psychiatry, 40, 240-246. interesting; however, such high placebo response rates are well-known with anxiety disorders. Bech.P., Kastrup.M. & Rafaelson.O.J. Studies on patients with panic disorder, for (1986) Mini-compendium of rating scales for example, have demonstrated placebo-response states of anxiety, depression, mania, rates in the range of 40 to 50% (Schweizer et schizophrenia with corresponding DSM-III al.,1993; Londborg et al.,1998). syndromes, <4cfa Psychiatrica Scandinavica, 73 This study had a few limitations. The (suppl. 326), 1-37. sample size was relatively small, which might have been the reason why certain obvious Cott.J., Misra.R. & Dev.S. (1994) Novel statistical trends escaped significance. The pharmacological profile of some psychoactive inclusion of multiple diagnoses precluded an medicinal extracts. Psychopharmacology identification of the value of Aswagandha in Bulletin, 30, 95. each; samples for individual diagnoses were too small to permit subgroup analyses. Handa.S.S. (1995) Plants and plant It is concluded that further investigation products for mental health. In : Decade of the of the anxiolytic properties of Withania somnifera Brain, (Eds.) Koslow.S.H., Murthy.R.S. &

300 ANXIOLYTIC EFFICACY OF ETHANOLIC EXTRACT OF WITHANIA SOMNIFERA

Coelho.G.V., pp 163-171. Rockville, MD : U.S. Psychopharmacol Bulletin, 22, 343-381. I Department of Health and Human Services. Londborg.P.D., Wolkow.R., Smith.W.T, lyengar,M.A. (1981) Study of Crude DuBoff.E., England,D., Ferguson.J., Rosenthal, Drugs. Manipal : Manipal Power Press, 1981. M. & Weise,C. (1998) in the treatment of panic disorder ; a multi-site, double-blind, Kessler.R.C, McGongale,K.A.; Zhao, placebo-controlled, fixed-dose investigation. S., Nelson,C.B., Hughes,M., Eshleman.S., British Journal of Psychiatry, 173, 54-60. Wittchen,H.U. & Kendler,K.S. (1994) Lifetime and 12-month prevalence of DSM-III R Mehta.A.K., Binkley.P., Gandhi.S.S. & psychiatric disorders in the United States. Ticku.M.K. (1991) Pharmacological effects of Archives of General Psychiatry, 51, 8-19. Withania somnifera root extract on GABA-A receptor complex. Indian Journal of Medical Kulkarni,S.K. & George,B. (1996) Research, 94, 312-315. Anticonvulsant action of Withania somnifera (Aswagandha) root extract against Satyavati.G.V. (1995) Leads from pentylenetetrazol induced kindling in mice. on medical plants acting on the Phytotherapy, 10, 447-449. . In : Decade of the Brain, (Eds.) Koslow, S.H., Murthy.R.S & Coelho.G.V., pp Kulkarni.S.K., Sharma.A., Verma.A. & 185-190. Rockville, MD : U.S. Department of Ticku.M.K. (1993) GABA receptor mediated Health and Human Services. anticonvulsant action of Withania somnifera root extract. Indian Drugs, 30, 305-312. Schweizer.E., Patterson,W., Rickels.K. & Rosenthal.M. (1993) Double-blind, placebo- Kulkarni.S.K., George,B. & Nayar,U. controlled study of a once-a-day, sustained- (1994) Amygdaloid kindling in rats : protective release preparation of for the effect of Withania somnifera (Aswagandha) root treatment of panic disorder. American Journal extract. Indian Drugs, 32, 37-49. of Psychiatry, 150, 1210-1215.

Levine.J. & Schooler.N. (1986) Stahl.S.M. (1998) Essential SAFTEE: A technique for the systematic Psychopharmacology. New Delhi : Cambridge assessment of side effects in clinical trials. University Press.

CHITTARANJAN ANDRADE'. MD. Additional Professoi & Head. ANITHAASWATH. MA. Research Assistant. Department of Psychopharmacology. S.K. CHATURVEDI. MD. Additional Professor. Department of Psychiatry.National Institute of Mental Health and Neurosciences. Bangalore-560 029. M SRINIVASA. DPM. Chief Consultant. Spandana Psychiatric Hospital, Bangalore-560010&R RAGURAM. MD, Additional Professor. Department of Psychiatry. National Institute of Mental Health and Neurosciences. Bangalore-560 029. * Correspondence

301